Research  /  Cannabis
WATCH

Cannabis & THC

THC · CBD · Endocannabinoid System · Modern Potency · Sleep · Mental Health · CUD · 25+ studies cited · April 2026

Cannabis is one of the most polarized health topics — celebrated as a panacea by some, demonized as gateway drug by others. The truth is genuinely complicated, and the evidence has shifted significantly in recent years as legalization has enabled better research and as products have become dramatically more potent than the cannabis of previous generations. This is not your dad's weed, and the old debates don't quite apply anymore.

What Cannabis Actually Is

Cannabis contains over 100 cannabinoids, but two matter most:

THC (Δ9-tetrahydrocannabinol) — the primary psychoactive compound. Binds CB1 receptors in the brain, producing the "high." Responsible for almost all the acute mental effects: euphoria, altered perception, anxiety/paranoia at higher doses, impaired short-term memory.

CBD (cannabidiol) — non-psychoactive. Doesn't produce a high. Acts as a partial agonist at the 5HT-1A serotonin receptor and a negative allosteric modulator of CB1. Has its own therapeutic profile, largely separate from THC.

These act on the endocannabinoid system — a network of CB1 (brain-dominant) and CB2 (immune-dominant) receptors and endogenous ligands (anandamide, 2-AG). This system regulates appetite, mood, sleep, pain, immune function, and memory. It's a real biological system, present in all vertebrates.

The Modern Potency Problem

This is the most underappreciated fact about cannabis today and the one that invalidates much of the older "weed is harmless" research.

Cannabis potency has increased 3- to 5-fold in the past two decades.

EraAverage THC %What It Was
1970s–80s~3–4%Mostly leaf and stem
1990s~6%Improved breeding
2000s~10%Sinsemilla becoming standard
2020s flower~17–25%Dispensary average
2020s concentrates60–95%Wax, shatter, dabs, vape carts

When studies from the 1980s say "moderate cannabis use isn't associated with X," they were studying a substantially different drug. The cannabis of 1985 and the cannabis of 2025 are pharmacologically related but functionally different products. Higher potency = greater risk of cannabis use disorder, cardiovascular events, mental health complications, and cognitive effects.

This is similar to comparing wine and grain alcohol — same active ingredient (ethanol), wildly different risk profiles.

Sleep — The Most Common Use Case, And It's Complicated

Probably the #1 reason adults use cannabis is "to help with sleep." The evidence here is more nuanced than people think.

What Acute Use Does to Sleep

Mixed Effects
  • Falls asleep faster — most users report shorter sleep onset latency. Real effect.
  • Subjectively feels deeper — many users report "best sleep ever" feelings.
  • REM sleep is suppressed — THC reduces time in REM sleep and prolongs REM latency. Documented in both older and newer studies.
  • N1 (light sleep) increases — sleep architecture shifts toward lighter sleep.
  • Sleep efficiency decreases with chronic use.

The REM Problem — Why Cannabis Sleep Isn't Real Sleep

Chronic Use Concern

REM sleep is when emotional processing, memory consolidation, and dream activity happen. Suppressing it night after night has consequences:

1. Emotional regulation suffers — REM is when your brain processes the day's emotions. Chronic suppression is associated with mood dysregulation.

2. Memory consolidation is impaired — particularly procedural and emotional memories.

3. REM rebound on cessation — when chronic users stop, they often experience intense, vivid, sometimes disturbing dreams as the brain tries to "catch up." This is one reason quitting feels so hard.

4. Withdrawal-induced insomnia — chronic users almost always have rebound insomnia when they stop, sometimes lasting weeks. This creates a self-perpetuating dependency: "I can't sleep without it" — but the reason you can't sleep without it is BECAUSE you've been using it.

What the Research Actually Shows

Meta-Analysis Evidence
StudyTypeKey Finding
Cannabis and Sleep Architecture (2024) Solid Systematic review & meta-analysis Frequent users showed longer REM latency, decreased sleep efficiency, and increased wake-after-sleep-onset. Effects vary by THC/CBD ratio, dose, and chronicity.
Chronic Cannabis & Sleep Architecture (SLEEP, 2024) Polysomnography cohort Objective THC metabolite measurement in urine was associated with prolonged REM latency and decreased REM proportion in a sleep-clinic cohort.
Oral Cannabinoids & Sleep — Pilot RCT (J Sleep Research, 2026) Randomized controlled trial THC/CBD significantly decreased REM time and increased REM latency compared to placebo.

Honest summary: Cannabis can help you fall asleep but degrades sleep quality, particularly REM. Chronic use creates a dependency cycle where withdrawal-induced insomnia makes quitting hard. If you're using cannabis primarily for sleep, you're trading one kind of sleep problem for another.

This connects directly to the cortisol/HPA axis — chronic sleep architecture disruption from any source dysregulates the HPA axis over time.

Mental Health — The Real Risks

This is where the evidence has gotten genuinely concerning, especially for adolescents and high-potency products.

Psychosis and Schizophrenia — The Strongest Signal

Causal Evidence

This is the strongest mental health signal in the cannabis literature, and it's getting clearer:

  • Mendelian randomization studies (which can establish causation by using genetic variants as natural randomizers) show cannabis exposure plays a causal role in schizophrenia development
  • The overall odds ratio for schizophrenia in cannabis users is approximately 2.88
  • Adolescent use roughly doubles the risk
  • High-potency cannabis (>10% THC) carries the highest risk
  • The mechanism likely involves THC disrupting endocannabinoid signaling during the critical adolescent brain development window

This isn't "reefer madness" propaganda. This is consistent epidemiological evidence supported by mechanistic studies and Mendelian randomization (which controls for the confounding that plagued earlier observational studies).

StudyTypeKey Finding
Cannabis & Schizophrenia — Mendelian Randomization (Molecular Psychiatry) Solid Mendelian randomization Cannabis exposure plays a causal role in schizophrenia development — genetic instrumental variables eliminate reverse causation.
Adolescent Cannabis & Schizophrenia — Updated SR (2022) Systematic review Adolescent use roughly doubles schizophrenia risk; high-potency products carry the highest risk.
Cannabis & Schizophrenia Causation Analysis (PMC, 2025) Causation analysis Consistent epidemiological evidence supported by mechanistic studies. OR ≈ 2.88 for cannabis users.

Key caveat: Schizophrenia has a baseline prevalence of ~1%. Doubling that to ~2% is a real and concerning increase but doesn't mean cannabis causes psychosis in most users. It does mean it's a serious risk factor, particularly for those with family history or genetic vulnerability, and especially in adolescents.

Depression — Bidirectional Relationship

Consistent Association

A 2025 systematic review and meta-analysis found:

  • Young cannabis users had 51% higher odds of depression (28% after adjustment for confounders)
  • The relationship may be bidirectional — depression can lead to cannabis use, and cannabis can worsen depression
  • Effect appears stronger in adolescents and heavy users

Anxiety — The Biphasic Paradox

Dose-Dependent

Cannabis has a complex anxiety relationship:

  • Acute low-dose THC — anxiety-reducing for many users
  • Acute high-dose THC — anxiety-PROVOKING (paranoia, panic attacks)
  • Chronic use — associated with higher baseline anxiety
  • CBD specifically — has some genuine anxiolytic evidence (see below)

The dose-response curve is biphasic — small amounts may calm, larger amounts agitate. Modern high-potency products make hitting the "wrong" dose much easier.

CBD — A Different Story

CBD without THC is a different drug. The evidence for CBD specifically is more favorable.

CBD for Anxiety — Substantial Effect Size

Meta-Analysis
StudyTypeKey Finding
CBD Therapeutic Potential in Anxiety — Meta-Analysis (2024) Solid Meta-analysis, 8 studies, n=316 Substantial significant effect size (Hedges' g = -0.92, 95% CI -1.80 to -0.04) for CBD on anxiety.
CBD in Anxiety Disorders — Systematic Review of RCTs (2024) Systematic review of RCTs Supports CBD as an anxiolytic option, with best evidence in specific anxiety disorders.
  • Epilepsy: FDA-approved for severe pediatric epilepsy (Epidiolex) — strongest cannabis-derived medication.
  • Pain: Mixed evidence for inflammatory and neuropathic pain.
  • Inflammation: Strong preclinical evidence, weaker clinical evidence.

CBD doesn't carry the psychosis risk, the REM suppression, or the dependence risk that THC does. If you're considering cannabis for anxiety or sleep, CBD-dominant or CBD-only products are a much better-supported option than THC-heavy products.

Cannabis Use Disorder — Yes, It's Real

A persistent myth in cannabis culture is "you can't get addicted to weed." This is false, and it's gotten less true as potency has increased.

Cannabis Use Disorder (CUD) is a recognized DSM-5 diagnosis. The criteria include:

Prevalence

PopulationCUD Rate
All people who have used cannabis~22% will develop CUD
All US adults (past year CUD)6.8% (~19.2 million)
Adults 18–25 (past year CUD)16.6%
21-year-olds (highest risk window)41.1%

These are not small numbers. Cannabis use disorder is more common than alcohol use disorder in young adults.

Withdrawal Is Real

Cannabis withdrawal symptoms are well-documented in DSM-5: irritability, anger, aggression, anxiety, sleep difficulty (insomnia, vivid dreams), decreased appetite or weight loss, restlessness, depressed mood, and physical symptoms (headache, sweating, abdominal pain, tremors).

Peaks at days 2–3, mostly resolves within 1–2 weeks. The sleep disturbance can persist longer.

Cardiovascular Effects — Emerging Concern

This is the area where the evidence has shifted most dramatically. Older research suggested cannabis was cardiovascularly neutral. Newer, larger studies suggest otherwise.

2025 Cardiovascular Meta-Analysis: MACE Risk

Emerging Evidence
StudyTypeKey Finding
Cardiovascular Risk & Cannabis — SR & Meta-Analysis (PubMed, 2025) Solid Systematic review & meta-analysis Positive associations between cannabis use and major adverse cardiovascular events (MACE) — including heart attack and stroke.

Mechanisms include:

  • Acute heart rate increase of 20–50% with smoking
  • Blood pressure changes (often increases acutely)
  • Endothelial dysfunction
  • Sympathetic nervous system activation
  • Possible arrhythmia risk in susceptible individuals

The association is strongest in: younger users (where baseline CV risk is otherwise low), heavy users, smoked products (vs edibles), and high-potency products.

This is an emerging concern, not a settled question, but the trend in the evidence is clearly toward more cardiovascular risk than previously appreciated.

All-Cause Mortality

A 2025 systematic review and meta-analysis — the first to directly examine cannabis use and all-cause mortality — found significant associations. The signal is weaker than alcohol but real, driven primarily by cardiovascular events, accidents, and mental health-related mortality.

Cognitive Effects

Pregnancy — Clear Harm

This is one of the few cannabis areas where the evidence is unambiguous. A 2024 meta-analysis on maternal and neonatal outcomes found cannabis use during pregnancy associated with:

Avoid Cannabis During Pregnancy

  • Low birth weight
  • Small for gestational age
  • Major congenital anomalies
  • Decreased head circumference
  • Decreased neonatal weight and birth length
  • Decreased gestational age
  • NICU admission
  • Perinatal mortality
  • Preterm delivery

Don't use cannabis during pregnancy. This is a clear "Avoid."

Legitimate Medical Uses

Cannabis isn't all risk. There are real medical applications with reasonable evidence.

ConditionEvidence QualityNotes
Severe pediatric epilepsy (Dravet, Lennox-Gastaut) Strong FDA-approved (Epidiolex/CBD), gold standard medical use
Chronic neuropathic pain Moderate Especially when other treatments have failed; living systematic reviews ongoing
Chemotherapy-induced nausea Strong Established medical use, including FDA-approved synthetic cannabinoids
MS-related spasticity Moderate THC/CBD combinations
Cancer cachexia / appetite stimulation Moderate Real effect, real value
PTSD Mixed Some benefit, some worsening — context-dependent
Anxiety (CBD specifically) Moderate See CBD meta-analysis above
Insomnia Weak — may be net negative Acute help, long-term problems
Depression Negative Associated with worsening, not improvement

Honest Assessment

What's well-established: Modern cannabis is 3–5x more potent than historical cannabis — this matters. Cannabis use disorder is real and affects ~22% of users. Adolescent use is associated with significant risks (psychosis, IQ effects, depression). Cannabis suppresses REM sleep and degrades sleep architecture with chronic use. Pregnancy use causes clear harm. Cannabis has legitimate medical uses for specific conditions.

What's genuinely debated: Adult occasional use risk profile. Long-term effects of moderate use. Whether benefits outweigh risks for specific use cases. Cardiovascular risk magnitude (emerging, increasing). Effects of CBD-only products at common doses.

What's overstated by critics: "Gateway drug" theory has weak evidence. "Permanent brain damage" claims for adult occasional use. Reefer madness-era claims of inevitable schizophrenia or addiction.

What's overstated by advocates: "It's completely safe." "It's better than alcohol" — depends on context, dose, individual. "Helps everyone sleep" — works acutely, degrades sleep architecture chronically. "No addictive potential" — flatly false. "Natural so it's safe" — modern concentrates aren't natural in any meaningful sense.

The practical position: This is a Watch on the evidence dashboard — meaning context-dependent, evolving evidence, and risk varies substantially by user, product, dose, and frequency. If you're going to use cannabis: avoid under age 25, avoid daily use, prefer lower potency, prefer CBD-dominant products for therapeutic effects, avoid long-term sleep use, don't drive, don't use during pregnancy, and be honest with yourself about dependence.

References & Primary Sources

Mortality & Cardiovascular

Relative Risk of All-Cause Mortality with Cannabis Use — SR & Meta-Analysis (PMC, 2025) Cardiovascular Risk Associated with Cannabis Use — SR & Meta-Analysis (PubMed, 2025) Solid Cannabis, Cannabinoids and Health: Risks and Benefits Review (Eur Arch Psychiatry Clin Neurosci, 2024)

Sleep

Cannabis and Sleep Architecture — SR & Meta-Analysis (PubMed, 2024) Solid Chronic Cannabis Use and Sleep Architecture — Sleep-Clinic Cohort (SLEEP, Oxford Academic, 2024) Cannabis Use Proximal to Sleep — Polysomnography (JCSM, 2024) Acute Effects of Oral Cannabinoids on Sleep — Pilot RCT (J Sleep Research, 2026)

Mental Health

Cannabis and Schizophrenia — Mendelian Randomization (Molecular Psychiatry) Solid Adolescent Cannabis Use and Schizophrenia — Updated Systematic Review (PubMed, 2022) Does Cannabis Use Contribute to Schizophrenia? Causation Analysis (PMC, 2025) Yale: Behind the Smoke — Cannabis and Schizophrenia (2024) Cannabis and Depression — Updated Meta-Analysis (PMC, 2025) Mental Health Risks in Youth — SR & Meta-Analysis (Addictive Behaviors, 2025) Cannabinoids for Mental Disorders & Substance Use — Lancet Psychiatry Meta-Analysis (2026)

Cannabis Use Disorder

Cannabis Use Disorder — StatPearls (NCBI) CDC: Understanding Your Risk for CUD Prevalence of CUD in People Who Use Cannabis — Systematic Review (PubMed, 2020) The Developmental Trajectory to CUD (American Journal of Psychiatry) DSM-5 CUD Construct Validity (PMC, 2022)

Pregnancy

Maternal and Neonatal Outcomes of Cannabis Use in Pregnancy — Meta-Analysis (PMC, 2024)

Chronic Pain

Living Systematic Review on Cannabis for Chronic Pain — 2025 Update (NCBI Bookshelf)

CBD Specifically

CBD in Anxiety Disorders — Systematic Review of RCTs (PMC, 2024) CBD Therapeutic Potential in Anxiety — Meta-Analysis (PubMed, 2024) Solid CBD for Pain — SR of Clinical and Preclinical Evidence (PMC, 2024)